Normal phase method not reproducible

[sdegrace]

We are running a normal phase method for a hydrophobic fluorescent small molecule analyte. We are observing major problems with reproducibility of the method and I was hoping to get some advice.

System: Waters Alliance 2695 with 2475 fluorescence detector (ex 290 nm, em 330 nm). We have two separate systems in two buildings which are running this method and experiencing the same problems. Both have been serviced by Waters recently, and the problems with the method are longer-standing.

Column: Silica, 250 mm x 4 mm, 5 um particle size.

Mobile phase: n-Hexane with 1% v/v Isopropyl Alcohol

Flow: 1 mL/min, isocratic

Sample compartment at 4 C, column at 25 C.

Injection volume is 10 uL. Sample diluent is n-hexane. The analyte is an additive in a fatty acid mixture. The FA mixture is directly dissolved in n-hexane to a concentration of 5 mg/mL, where the analyte is therefore at a nominal concentration of about 15 ug/mL. The standards are the analyte directly dissolved in n-hexane at about the same concentration. Analysis is w/w% assay against external standard. Detector response is very strong and we are nowhere near the LOQ.

We advised Waters on install of the systems of the intention to use with normal phase techniques, and the systems should be set up appropriately.

What happens is that on a given run, the standard area counts will often creep higher, causing system suitability to fail. We never see carryover into the blanks, though. Similar reproducibility problems can be seen with the sample preparations.

We have considered equilibration of the system and purging the injector a potential root cause, and very through purging with IPA and then mobile phase prior to a run seems to help in some but not all cases. We have wondered about evaporation, but the sample compartment is kept cool, and we think we have to use the slitted vials with the Alliance system.

We're wondering if anyone else has seen similar issues and might be able to offer some advice.

Thanks,

Stephen

[JGK]

Have you tried using a -CN or -NH2 phase which are better than silica for equilibration. Alternatively look at HILIC stationary phases

[sdegrace]

Have you tried using a -CN or -NH2 phase which are better than silica for equilibration. Alternatively look at HILIC stationary phases

If I had latitude to seriously alter the method I would be interested in that, but I received the method as-is and I don't have a ton of latitude to change it. HILIC might be close enough to straight silica that I could look at it but I feel like it is something to do with the system and not the column. Very similar method with UV detection worked well on an older Varian system so I am suspicious that something is happening with the Alliance that I'm not understanding, or that it relates to the detection somehow.

[HPLCaddict]

I don't think that this is a equillibration issue, as this would rather lead to irreproducible retention times, not increasing peak areas.
My first thought would be evaporation of the sample solvent - hexane is very volatile...

[sdegrace]

I don't think that this is a equillibration issue, as this would rather lead to irreproducible retention times, not increasing peak areas.
My first thought would be evaporation of the sample solvent - hexane is very volatile...

I wonder if our problem is the slitted septa used with the vials for the Alliance autosampler. I'll contact Waters and see if they can recommend something better for more volatile substances.